Why are many electrolytics not specified ESR?

[throbscottle]

Take Panasonic FC series capacitors, widely available and inexpensive. Or Rubycon ZL series, so Pana doesn't feel lonely. Neither of them are specifying ESR for these in their datasheets. Why is this?

Thanks
Me

 

[JimB]

Look what I just found:

 

[throbscottle]

It's not the same thing. Take this excerpt from Cornell Dublier's guide to electrolytic capacitors https://www.cde.com/resources/catalogs/AEappGUIDE.pdf :

ESR Frequency Characteristics
Like DF, the ESR varies with frequency. Rewriting the DF
equation above, ESR can be modeled as below:

ESR = 10,000(DF lf )/(2πfC) + ESR hf
Expressing the ideas in ESR terms, at low frequencies the
ESR declines steadily with increasing frequency and crosses
over to constant ESR at a frequency inversely proportional to
capacitance. This crossover is typically below 10 kHz. The ESR
of high-capacitance capacitors changes little with increasing
frequency because high-capacitance causes them to have low
crossover frequencies. The ESR hf ranges from 0.002 Ω for large,
screw-terminal capacitors to 10 Ω for miniature devices.
IMPEDANCE (Z)
For aluminum electrolytic capacitors impedance is actually
impedance magnitude. It is the ratio of voltage to current at
a given frequency and is related to the capacitor’s capacitance,
ESR and series inductance as follows:
Z = [(ESR)2 + (1/(2πfC) – 2πfLs)2] 1⁄2
Where Z is impedance in Ω, ESR is equivalent series resistance in
Ω, f is frequency in Hz, C is capacitance in F and Ls is equivalent
series inductance in H.

So they are saying ESR is a part of impedance.

 

[Nigel Goodwin]

It's not the same thing.

Yes it is, that's just how that particular datasheet expresses it - as to which method might be 'right' or 'wrong', I would say it simply depends on how you look at it

For example ESR - specifies 'resistance' actually in the name - but as this is an AC measurement (normally at 100KHz) shouldn't it more correctly be called 'impedance'?.

So don't overcomplicate things for yourself

 

[throbscottle]

Hmmmm grumble grumble grumble...

 

[ronsimpson]

So don't overcomplicate things for yourself

Capacitance
ESR
ESL
dielectric absorption, soakage, or memory effect
All the above is effected by temperature and age.

Some one could write yet another book on the simple little capacitor.

 

[Nigel Goodwin]

Capacitance
ESR
ESL
dielectric absorption, soakage, or memory effect
All the above is effected by temperature and age.

Some one could write yet another book on the simple little capacitor.

Yes, and how pointless would it be

 

[throbscottle]

I had to ask because I'm on the point of actually spending some money and want to make level comparisons. So I'll get the Pana FC series ones and if they prove unsuitable I will have to hunt you down and throw capacitors at you...

 

[Nigel Goodwin]

I try and stock only Panasonic and Rubycon 105 degree capacitors (and source them from RS Components), the major part of servicing work in the recent past was replacing high ESR electrolytics.

However, servicing is pretty well dead now

 

[Tony Stewart]

Take Panasonic FC series capacitors, widely available and inexpensive. Or Rubycon ZL series, so Pana doesn't feel lonely. Neither of them are specifying ESR for these in their datasheets. Why is this?

Thanks
Me

All electrolytics are defined by either low ESR with value, Ripple Current at 120Hz at T max or Trise and Dissipation Factor , D at 120Hz.

This is the standard.

ESR is defined from each of the above or implied, however, in my experience, if Ripple current only is given, it may not be as low ESR as those explicitly stated and guaranteed when new. However depending on the application, Cap life can be extended by adding external Rs to minimize the temperature rise from low % ripple designs with equally low Average/Peak ratios, which means massive peaks for <5% ripple or x20 Avg.

These are design tradeoffs for efficiency and life expectancy with levels of ripple and unregulated errors.

Mil Std Handbook 217 defines e-Cap reliability with an accelerated failure rate based on external circuit ESR which affects ripple peaks and I^2ESR dissipation power rise times thermal R insulation factor that results in a certain temperature rise that causes aging.

 

[throbscottle]

OK then, so my general plan is to put 3 x 680uF in parallel because I think the ESR will be lower than a singe 2200uF, however it still has to be higher than 0.03Ω (output cap for a buck converter), since it works out a bit cheaper - hence the original question.
The lifetime and temperature stuff is a bit over my head I'm afraid

 

[ronsimpson]

output cap for a buck converter

The ripple current may be lower than you think. It is very dependent on the inductor used. It is probably the same as the current in the coil.
What I am trying to say is that if the inductance is increased the capacitor current is reduced.

 

[chemelec]

According to my Secore Capacitor/Inductor Analyzer, they have This Documentation.

But I Find the Smaller SIZE Capacitors are Usually HIGHER ESR and Worse than these Recommended Values.

So that Chart in Post #2 Makes sense.

 

[Tony Stewart]

OK then, so my general plan is to put 3 x 680uF in parallel because I think the ESR will be lower than a singe 2200uF, however it still has to be higher than 0.03Ω (output cap for a buck converter), since it works out a bit cheaper - hence the original question.
The lifetime and temperature stuff is a bit over my head I'm afraid

Don't guess, as I said it is all specified and you never indicated your selection part number.

ESR can be defined from Dissipation Factor or tan Delta and is given for the series you specified.

 

[throbscottle]

Looks like I'll have to actually work through the formulas in that informative little book then (post #3).
The 680uF I was looking at is part no. EEUFC1H681, though comparing the impedance of 3 in parallel with a single 2200uF from the same series, it's about the same. (and according to chemelc's chart it's worse) The combined ripple current is around 6A though, compared to around 3A for the single 2200uF.

 

[Tony Stewart]

EEU-FC1H222
Ripple Current 3100mA @100kHz @105'C
ESR 23 mOhm . . . . . . . @100kHz @20'C
Lead Spacing 0.295" (7.50mm)
Size / Dimension 0.709" Dia (18.00mm)
Height - Seated (Max) 1.398" (35.50mm)

EEU-FC1H681L x3 <<< best bet.
Ripple Current 1.661A
ESR 40 mOhm . . . . . . . @100kHz @20'C
Lead Spacing 0.197" (5.00mm)
Size / Dimension 0.492" Dia (12.50mm)
Height - Seated (Max) 1.181" (30.00mm)

For better redundancy and reliability, Ripple Current and ESR 3x EEU-FC1H681L is much better at only slightly more cost (/1k) but at 25% more board space.and almost twice the height ( free ?)

The size matters in terms of ripple current. Although I might consider 3x 820uF and only try 2 for stability reasons. (ref spec)

Switch ESR is ~ 170mOhm or 340mOhm @ 50% d.c.
Output Ripple Voltage = (ΔI IND ) (ESR of C OUT )

Thus worst case 3A* 13mOhm = 42mV

 

[throbscottle]

Thanks I'll give them a go and see how it goes...